AWS Community Cameroon, Knowledge Bases and generative AI

As a Technology Strategist for Developer Advocacy efforts in the Sub-Saharan Africa region, one of my most favorite parts of my role is spending in-person time with members of the AWS community. For me, these are invaluable times of learning, where I get to learn about all that the community is currently building on AWS, and about any challenges that we as AWS can work together with them towards resolving.

Last week I spent some in-person time with the amazing community from Cameroon, specifically the AWS technical content creators of Cameroon. We discussed about their AWS content creation processes, and I got to enjoy a lot of inspiring stories about how many started their AWS journeys.

I look forward to the next few months of working closely with this group, as we look into various strategies to support their content creation efforts, and have them raise the bar on what's already awesome content being created by this group.

I thank the leaders of this community for the invite Nkwenti Fon Nkwenti Ngwa Bandolo Bobga Cyril Paula Ali Wakabi Chi Che. Ndimofor Ateh Rosius , and I especially thank everyone who took time off their busy schedules, to attend this session.

Now, onto what I've been learning in AWS lately....

Knowledge bases and generative AI

Generative AI is transforming knowledge work, enhancing decision-making, increasing productivity, and driving digital transformation. With business operations that are changing all the time, it can be challenging for us as colleagues and our customers to keep up with these changing operations, and to keep up with the information that's influencing these changes. At most times, this information is contained in multiple documents, policies, processes, or that one colleague who keeps promising to document it all one day. This information is usually unstructured, scattered in various repositories, making it a challenge to find and consume this information.

Organizations are usually challenged in terms of the best way to make use of their proprietary knowledge using generative AI. While pre-trained models offer remarkable capabilities, they lack the ability to express views regarding internal knowledge bases as they were not trained using this internal information. Now organizations can bridge the gap between this proprietary knowledge and the offerings of generative AI.

Retrieval-Augmented Generation (RAG) optimizes the output of large language models (LLMs) by referencing your proprietary knowledge base before generating a response. RAG extends the already powerful capabilities of LLMs to your organization's internal knowledge base, and the cool part is that all without the need to retrain the model. This makes it a cost-effective, time saving approach to improving LLMs' output thus making it relevant, accurate, and useful in various contexts.

Use cases of knowledge bases in generative AI

Ever since the introduction of generative AI, many have been looking for value-adding use cases of this technology. Here are a few that I've been looking into:

• Smart Q&A applications - an educational Q&A applications that interrogates your internal documents, and responds with answers that have context. In Amazon Bedrock I'm building (and testing) a knowledge base whose data source is the Wildlife in South Africa Wikipage. It's meant to be a knowledge base for a fun application that will be interrogated to learn more about South Africa and it's wildlife.

• Troubleshooting guides and tutorials - Build context-rich generative AI-powered knowledge bases that let your customers to find solutions independently.
• Unstructured content - It's challenging to organize and analyze unstructured content. This is because of its varied formats and lack of standardized structure. Knowledge bases can provide you with the context needed to analyze this unstructured content.
• Use across channels - Extend your knowledge base by using it on applications across various platforms such as live chats and social messaging.

How will you be using knowledge bases in generative AI to enhance the experiences of your colleagues and your customers?

References:

1. Knowledge Bases for Amazon Bedrock
2. How Generative AI is revolutionizing Knowledge Management
3. Accenture creates a Knowledge Assist solution using generative AI services on AWS